Sound Change Quickie Thread

[WeepingElf]

Is k͡p → kʷ attested? If so, could PIE labiovelars have been coarticulated stops at some earlier point?

It's not impossible, but /k͡p/ seems out of place in Eurasia; it is pretty much an African thing. In my personal opinion, PIE *kʷ etc. are from plain velars next to rounded vowels before the vowel system of Pre-PIE collapsed, perhaps due to a Para-NW-Caucasian substratum (the same kind of change seems to have happened in NW Caucasian, too).

[Travis B.]

Is k͡p → kʷ attested? If so, could PIE labiovelars have been coarticulated stops at some earlier point?

It's not impossible, but /k͡p/ seems out of place in Eurasia; it is pretty much an African thing. In my personal opinion, PIE *kʷ etc. are from plain velars next to rounded vowels before the vowel system of Pre-PIE collapsed, perhaps due to a Para-NW-Caucasian substratum (the same kind of change seems to have happened in NW Caucasian, too).

That hypothesis implies that PIE *kʷ etc. should pattern with PIE *u~*w -- do we have any positive evidence for this?

[WeepingElf]

Is k͡p → kʷ attested? If so, could PIE labiovelars have been coarticulated stops at some earlier point?

It's not impossible, but /k͡p/ seems out of place in Eurasia; it is pretty much an African thing. In my personal opinion, PIE *kʷ etc. are from plain velars next to rounded vowels before the vowel system of Pre-PIE collapsed, perhaps due to a Para-NW-Caucasian substratum (the same kind of change seems to have happened in NW Caucasian, too).

That hypothesis implies that PIE *kʷ etc. should pattern with PIE *u~*w -- do we have any positive evidence for this?

The idea is that most (if not all) vowels collapsed into a single vowel phoneme which then developed the various ablaut grades, and the old vowel features [+front] and [+round] were transferred to adjacent velars, if any.

[Travis B.]

It's not impossible, but /k͡p/ seems out of place in Eurasia; it is pretty much an African thing. In my personal opinion, PIE *kʷ etc. are from plain velars next to rounded vowels before the vowel system of Pre-PIE collapsed, perhaps due to a Para-NW-Caucasian substratum (the same kind of change seems to have happened in NW Caucasian, too).

That hypothesis implies that PIE *kʷ etc. should pattern with PIE *u~*w -- do we have any positive evidence for this?

The idea is that most (if not all) vowels collapsed into a single vowel phoneme which then developed the various ablaut grades, and the old vowel features [+front] and [+round] were transferred to adjacent velars, if any.

You would expect PIE *w to also condition labialization if that were true (it would be odd if a *u conditioned labialization but *w did not).

[ophois]

Is it possible for /s/ + voiced plosive to become a fricative without /s/ devoicing the following sound?

[/ˌnɐˈɾɛndɚduːd/]

Is it possible for /s/ + voiced plosive to become a fricative without /s/ devoicing the following sound?

I'd say it's unlikely, but still certainly possible, especially if your language already tolerates mixing voiced and voiceless consonants in clusters.

[Travis B.]

Is it possible for /s/ + voiced plosive to become a fricative without /s/ devoicing the following sound?

I'd say it's unlikely, but still certainly possible, especially if your language already tolerates mixing voiced and voiceless consonants in clusters.

Actually my main objection to this scenario is that fricatives at the beginnings of clusters tend to prevent frication (but not necessary affrication) of a following stop, as shown by how fricative + voiceless plosive clusters in PIE survived both the Germanic consonant shift and the High German consonant shift.

[malloc]

Is it possible for /s/ + voiced plosive to become a fricative without /s/ devoicing the following sound?

For that matter, would the /s/ voice in this scenario? Some form of voicing assimilation is quite likely.

[ophois]

Is it possible for /s/ + voiced plosive to become a fricative without /s/ devoicing the following sound?

I'd say it's unlikely, but still certainly possible, especially if your language already tolerates mixing voiced and voiceless consonants in clusters.

Actually my main objection to this scenario is that fricatives at the beginnings of clusters tend to prevent frication (but not necessary affrication) of a following stop, as shown by how fricative + voiceless plosive clusters in PIE survived both the Germanic consonant shift and the High German consonant shift.

Albanian had medial /sp sd sk/ > /f θ h/ (/st/ became /ʃt/) so it's not out of the question for the sound change to happen with voiceless consonants. Brythonic, which I'm heavily cribbing from, also had word-final /s/ debuccalise, then geminate a following word-initial stop, with geminates then becoming fricatives.

The plan is for words originally ending in *-l/r/s to all condition a single mutation type on the following consonant. I could have *sb sd sg simply become /b d g/, and then have them either correspond to fricatives in that environment by analogy with words ending in liquids or have them undergo lenition like the regular voiced stops did in the same environment.

Is it possible for /s/ + voiced plosive to become a fricative without /s/ devoicing the following sound?

For that matter, would the /s/ voice in this scenario? Some form of voicing assimilation is quite likely.

The /s/ would be ultimately deleted, but whether it voices mid-way through the process is dependent on how plausible it makes my desired end result.

[Qwynegold]

How plausible are these sound changes?

mh, nh, ŋh → m̥, n̥, ŋ̊ → p, t, k

The idea is that /h/ devoices preceding consonants, even in a preceding word. /h/ is then elided, but <h> is kept in writing. This would create consonant mutation at the end of words, when the next word is one that used to begin with /h/. If the above sound changes are plausible, then is this consonant mutation idea plausible? If it's implausible for all h- words to cause the mutation, then what about just a few really common words causing it?

[Travis B.]

How plausible are these sound changes?

mh, nh, ŋh → m̥, n̥, ŋ̊ → p, t, k

The idea is that /h/ devoices preceding consonants, even in a preceding word. /h/ is then elided, but <h> is kept in writing. This would create consonant mutation at the end of words, when the next word is one that used to begin with /h/. If the above sound changes are plausible, then is this consonant mutation idea plausible? If it's implausible for all h- words to cause the mutation, then what about just a few really common words causing it?

I don't see anything really implausible about these sound changes myself. Both the first set of changes, and the second set of changes, are perfectly plausible by themselves, and I don't see any reason against the two sets of changes happening one after another.

[Imralu]

Is k͡p → kʷ attested? If so, could PIE labiovelars have been coarticulated stops at some earlier point?

It's not impossible, but /k͡p/ seems out of place in Eurasia; it is pretty much an African thing.

Vietnamese has them, as do quite a few languages in Melanesia, particularly in Vanuatu.

[Travis B.]

Is k͡p → kʷ attested? If so, could PIE labiovelars have been coarticulated stops at some earlier point?

It's not impossible, but /k͡p/ seems out of place in Eurasia; it is pretty much an African thing.

Vietnamese has them, as do quite a few languages in Melanesia, particularly in Vanuatu.

I wouldn't exactly call Melanesia "Eurasia" myself.

[Nortaneous]

Is it possible for /s/ + voiced plosive to become a fricative without /s/ devoicing the following sound?

I'd say it's unlikely, but still certainly possible, especially if your language already tolerates mixing voiced and voiceless consonants in clusters.

Actually my main objection to this scenario is that fricatives at the beginnings of clusters tend to prevent frication (but not necessary affrication) of a following stop, as shown by how fricative + voiceless plosive clusters in PIE survived both the Germanic consonant shift and the High German consonant shift.

How strong is this tendency outside Germanic presigmatized stops?

[Travis B.]

I'd say it's unlikely, but still certainly possible, especially if your language already tolerates mixing voiced and voiceless consonants in clusters.

Actually my main objection to this scenario is that fricatives at the beginnings of clusters tend to prevent frication (but not necessary affrication) of a following stop, as shown by how fricative + voiceless plosive clusters in PIE survived both the Germanic consonant shift and the High German consonant shift.

How strong is this tendency outside Germanic presigmatized stops?

This wasn't limited to presigmatized stops -- /ft/ (or shall I say /ɸt/?) and /xt/ clusters were also left intact by the High German consonant shift when otherwise /t/ affricated.

[Imralu]

It's not impossible, but /k͡p/ seems out of place in Eurasia; it is pretty much an African thing.

Vietnamese has them, as do quite a few languages in Melanesia, particularly in Vanuatu.

I wouldn't exactly call Melanesia "Eurasia" myself.

I wouldn't call it "Africa".

[Qwynegold]

I don't see anything really implausible about these sound changes myself. Both the first set of changes, and the second set of changes, are perfectly plausible by themselves, and I don't see any reason against the two sets of changes happening one after another.

Thanks! I was worried about the second step, because voiceless nasals don't sound similar to plosives to me. But it should be okay, since nasals and plosives are both stop consonants.

[Travis B.]

I don't see anything really implausible about these sound changes myself. Both the first set of changes, and the second set of changes, are perfectly plausible by themselves, and I don't see any reason against the two sets of changes happening one after another.

Thanks! I was worried about the second step, because voiceless nasals don't sound similar to plosives to me. But it should be okay, since nasals and plosives are both stop consonants.

Remember that in many languages voiced nasals and voiced plosives are closely related. If you want to more strongly justify the changes you detailed, I would do something like have an intermediate state of m̥, n̥, ŋ̊ > m̥p, n̥t, ŋ̊k and potentially transfer the nasalization to the preceding vowel (my native dialect of English has a productive sound change where coda mp, nt, ŋk > ʔp, ʔ(t), ʔk (and also when followed by a sibilant in the same coda) with nasalization of the preceding vowel, but the nasal consonant is preserved if there is a syllable boundary between it and the plosive).

[chelsea]

Any precedent for hp ht hk -> p' t' k' ?

What about ʔp ʔt ʔk -> p' t' k' ?

[Man in Space]

Any precedent for hp ht hk -> p' t' k' ?

What about ʔp ʔt ʔk -> p' t' k' ?

I could see **hp **ht **hk (> **pp **tt **kk) > *ʔp *ʔt *ʔk > p’ t’ k’. Directly from preaspiration to ejective would seem to be a stretch but if there’s a sequence in the sound changes it cound get there with a pit stop or two along the way.